This is a renewal of a K24 Career Investigator Award on sleep disturbances in alcohol-dependent patients. Insomnia is common, frequently persistent in newly recovering patients, and associated with relapse. In the first 5 years, the investigator acquired knowledge about sleep medicine, sleep research methods, traditional sleep stage scoring, and pharmacotherapy trial methods. He then designed and conducted a randomized, controlled pilot trial to study night-time administration of the sedating anticonvulsant drug, gabapentin, for its effects on both sleep and relapse to alcohol. This work produced the first evidence from a randomized controlled trial that gabapentin was superior to placebo for reducing time to relapse. Although robust effects on traditional measures of subjective and objective sleep were not found, analysis of sleep microarchitecture revealed that gabapentin increased delta (slow frequency) activity during rapid eye movement sleep, the period of sleep during which the cortex is paradoxically activated. Thus, gabapentin may have exerted its effect on daytime abstinence by slowing cortical hyperactivity during sleep. The current proposal builds on this work by investigating potential mechanisms for sleep disturbance in alcohol-dependent patients that could be targeted by medications. A novel sleep laboratory paradigm is described (and will be refined) that can inform the field about mechanisms of alcohol-related sleep disturbances as well as screen for medications that improve sleep. To demonstrate proof-of-concept, two medications with known potential for treating either alcohol dependence or insomnia (gabapentin, trazodone) will be studied for their short-term effects on disregulated sleep in alcohol-dependent patients. Training will comprise 10-15% effort for 5 years including mostly time in the sleep lab, but also local and national conferences on the neurobiology of sleep and alcoholism, and selected readings. Another 5-10% effort is for training new investigators supported by T32 and other fellowships. With Drs. Armitage and Hoffmann, the candidate will acquire new knowledge and skills about quantitative sleep EEG analysis and interpretation, as well as methods for measuring and analyzing homeostatic sleep drive and circadian rhythm regulation. He will apply these methods to alcohol dependent patients, an understudied population that may benefit clinically from a more sophisticated understanding of their sleep problems.